A decade later, the light rail line completed, university researchers have shown that the project did not, in fact, reduce congestion.
But forget the conclusion, says study co-author Sandip Chakrabarti — the fact that this kind of research is even possible is a big deal by itself. In 2005, nobody had access to the kinds of tools that could begin to approach an answer to that question. They could predict and they could model, but they couldn’t do anything close to what the University of Southern California did with the report it released in September.
The study, published in the Journal of Planning Education and Research, pulls together data from sensors across a huge chunk of Los Angeles — car speed, bus ridership, light rail ridership, peak congestion and more. It’s highly granular, allowing the researchers to explore traffic impacts on I-10 and nearby arterials before and after the light rail line opened for business, and then compare the before-and-after statistics to changes in control areas in different parts of the city.
That level of detail in an impact analysis of a major public transit project is something Chakrabarti and his fellow researchers could not find in the world of planning research.
“This is the first time where we were able to empirically show this using data,” Chakrabarti said.
The closest example he could find was a 2012 study looking at congestion on highways in Denver following the city’s light rail work. However, that study only examined freeway traffic and didn’t control for outside influences on the system like land use changes. The study suggested congestion reduction associated with the city’s light rail, something that was missing when Chakrabarti’s team finished its analysis of Los Angeles’s Expo Line.
The takeaway, Chakrabarti said, should be that more public officials should be using in-depth, granular, controlled analysis when trying to assess the true impact of transportation projects. That kind of work could change the conversation on the need for projects to something a little more accurate — while freeway congestion relief might be a common justification for light rail construction, Chakrabarti found that the Expo Line offered up some alternative benefits.
“People do believe that it would magically solve congestion, and we showed that it’s difficult and it hasn’t reduced congestion. But on the other hand it’s possible that it has restricted any deterioration in traffic conditions.”
Among the benefits the team did find: some congestion relief on arterials leading off of I-10, as well as ridership on public transit along the corridor — including buses — increasing overall after the light rail started operating. Public transit use is in itself a benefit, he argued.
“These kinds of investments … sort of create this alternative, this attractive alternative to car travel,” he said. “A lot of people who are sick and tired of getting stuck in traffic, it does give them an alternative.”
The increase in ridership basically means the corridor is serving more people, and that’s good for the city.
“It’s helped [people] access jobs and critical services more efficiently, and that’s a great thing,” he said.
It's not the first major project that has failed to tackle the difficult problem of congestion. As The Los Angeles Times has reported, a four-year, $1 billion expansion of Los Angeles' 405 Freeway failed to improve travel times.
Those kinds of insights are a product of very recent developments on the timescale of urban development — for instance, a national traffic data-sharing and analytics service run out of the University of Maryland has been operating since 2010. The program, called the Regional Integrated Transportation Information System (RITIS), allows something close to the USC analysis to users in a matter of minutes. Michael Pack, director of the university’s RITIS-hosting Center for Advanced Transportation Technology Laboratory, said that it does not, at the moment, have access to information about transit ridership like the USC study did.
But in many states, the program can call up much of the data needed to understand how a light rail has impacted congestion with the click of a button.
“It’s … changing the way people do business in that this analysis used to take months and months and months, with lots of people working on it, and now you and I can sit down in front of a computer and do it in 10 minutes,” Pack said. “It’s badass.”
On top of impact analysis, the tool also gives planners the ability to target their investments where they’re most needed.
“Those tools are used by a whole lot of DOTs to figure out, OK — where’s the worst congestion in my state and how expensive is that congestion, what are the speeds, and then if I did do a project, what type of impact would I have on the system? Would I reduce the cost of the congestion by $10 million a week or would I push the traffic down the road a little bit?” he said.
That could help save time and effort on projects that don’t offer a whole lot of benefit to the traveling public. It could also help break up models where priorities are driven by less-than-ideal sources.
“In the past, you had this state and local [workforce] and they would spend money on these projects sometimes — not all the time, but sometimes — based on whichever congressman was complaining the most about traffic in their district or based on people calling and complaining,” Pack said. “But now we have real data.”
The insights can also save states a lot of money, he said. State officials have to buy into the system — necessary to support RITIS’s model of buying data from private companies like INRIX and TomTom — but once they do, it offers them the kind of data that might otherwise eat up the budget of a typical research project.
“You can spend up to 50 cents on the dollar trying to locate data, buy data, or [interpret it] so you can actually use it,” he said.
After five years of operation, the system still has room to grow. According to Pack, only about 20 states and 5,000 people use the system. Knowledge of RITIS mostly spreads by word of mouth, and it can’t offer the same data for every state.
But Chakrabarti said he hopes to see the use of in-depth traffic analytics grow across the country.
“Increasingly, our transportation networks are being instrumented," he said, "meaning there are sensors everywhere — on freeways, on arterials, on surface streets, on transit vehicles."
